262 BULLETIN OF THE 



generally too small and rusty for this. The trap fragments in the 

 sandstone vary from small grains up to pieces one to three inches in 

 diameter; some are dense, some vesicular. This locality is about one 

 fourth of a mile north of a strong gap in the posterior ridge, and is 

 marked by a five-foot boulder of light granite. Crossing the marshy 

 stream, a small outcrop of ordinary normal sandstone is seen on the 

 opposite slope ; similar outcrops are found one fourth of a mile farther 

 northeast, higher up on the face of the posterior ridge ; and finally, just 

 below its trap, several feet of well metamorphosed sandstone appear, 

 gray in color and quartzitic in texture. The trap hei'e shows nothing 

 peculiar. Crossing over the ridge and descending a quarter of the way 

 to the river, we leave the trap and come to numerous outcrops of con- 

 glomeratic sandstone, darker in color than usual, and containing plen- 

 tiful scraps of trap, many of them vesicular, and in size up to three 

 inches. The first of these outcrops is not more than fifteen or twenty 

 feet over the trap ; many others appear at a greater distance, even on 

 the railroad a little below Smith's Ferrj' Station. It was there that 

 I first saw them in 1877, in company with Professor N. S. Shaler and 

 Mr. J. S. Diller. 



The upper contact was not found among these outcrops, but it is 

 excellently shown in Delany's Quarry, on the railroad and river bank, 

 not quite half-way from Smith's Ferry to Holyoke. (From the quarry 

 to Holyoke Station is about an hour's tiresome walk along the hot, sandy 

 track.) Here the rock is freshly worked, and the upper surface of the 

 trap is well shown to be very amygdaloidal, uneven, and knobby, as a 

 lava flow might be, and upon it lies the fine dull dark reddish muddy 

 shale, fitting closely to the trap and filling up its inequalities, so that 

 the sandstone a few feet higher is evenly bedded (see fig. 37). This 

 contact shale is as soft as ordinary shale at a distance from the trap ; it 

 is easily scratched to powder with a knife. On some trap faces patches 

 of similar shale appear to be included in the trap, but as they also are 

 not metamorphosed, I consider them to be muddy fillings of cavities 

 near the rough old lava surface, reached by passages not now exposed 

 to view ; where their bedding shows, it is about parallel to that of the 

 sandstone above. There was no appearance of branching intrusion into 

 the sandstone, or of breaking across its layers. I looked carefully at a 

 great number of freshly exposed amygdules here and elsewhere, in hopes 

 of finding some of them banded like those at Brighton, Mass. (see Bos- 

 ton Soc. Nat. Hist. Proc, XX., 1880, 42G), but was always unsuccessful. 

 Ten or fifteen feet of sandstone over the trap are well shown in the 



